A new study intended to help clinicians has indicated that artificial intelligence can predict suicide attempts up to two years in advance, with research revealing an 80 – 90% accuracy rate that increases the closer to the date of a suicide attempt.
The study, led by Florida State University researcher Jessica Ribeiro, uses artificial intelligence and machine learning algorithms in order to indicate how likely a person is to attempt suicide in the near future. Using 2 million electronic health records from Tennessee patients, these algorithms assessed a variety of factors in order to determine which patients were most at risk.
Also: North Korea Allows Citizens to Use Cell Phones So It Can Spy On Them
The largest research study of its kind, the wealth of data made available to Ribeiro and her team allowed them to comb through patient history (which was kept anonymous), finding those who had attempted suicide. Comparisons were then made between these patients’ health records, with the algorithms eventually being able to “learn” which combination of factors increased the risk of a patient attempting suicide.
The study was the largest of its kind. (Image Credit: Richard Wareham Fotografie / Getty Images)
The information gathered from this study led to Ribeiro and her team eventually being able to identify those who were at risk of suicide with great accuracy, which will be a vital tool for clinicians in the future. As Ribeiro explained to Florida State University, the machine learning algorithm developed by her team could lead to clinicians being sent suicide risk assessments, similar to the cardiovascular risk score they currently receive. This would enable them to be alerted when a patient was flagged as being a high suicide risk, therefore allowing them to pursue the appropriate course of action in order to help them.
With nearly 45,000 Americans dying by suicide each year, previous attempts at trying to find a correlation between suicide and other health issues was fruitless. Studies showed that depression, anxiety and substance abuse problems were not accurate indicators of a patient’s suicidal behavior, and that “60-90 percent of people who had died by suicide had visited their medical provider within the past year and the clinician never saw it coming.”
The study, titled “Predicting Risk of Suicide Attempts over Time through Machine Learning,” will be published by the journal Clinical Psychological Science. Ribeiro hopes that her team’s breakthrough will help prevent suicide on a larger scale, with the artificial intelligence algorithms granting clinicians the ability to treat a wider portion of the population dealing with suicidal thoughts.
